Email: ###@bu.edu (cmav instead of hashes)
Address:
Computer Science Department, MCS 135
Boston University
111 Cummington Mall
Boston, MA 02215
C. Mavroforakis, I. Valera and M. Gomez Rodriguez.
Modeling the Dynamics of Learning Activity on the Web
International World Wide Web Conference, WWW 2017 [code]
C. Mavroforakis, D. Erdos, M. Crovella and E. Terzi.
Active Positive-Definite Matrix Completion
SIAM International Conference on Data Mining, SDM 2017 [supplementary material]
C. Mavroforakis, I. Valera and M. Gomez Rodriguez.
Hierarchical Dirichlet Hawkes Process for Modeling the Dynamics of Online Learning Activity
Workshop on Networks in the Social and Information Sciences, NIPS 2015
C. Mavroforakis, M. Mathioudakis and A. Gionis.
Absorbing Random-Walk Centrality: Theory and Algorithms
ICDM 2015 [slides] [code]
C. Mavroforakis, N. Chenette, A. O'Neill, G. Kollios, and R. Canetti.
Modular Order-Preserving Encryption, Revisited
SIGMOD 2015
C. Mavroforakis, R. Garcia-Lebron, I. Koutis, and E. Terzi.
Spanning Edge Centrality: Large-scale Computation and Applications
International World Wide Web Conference, WWW 2015 [code]
M. Vlachos, F. Fusco, C. Mavroforakis, A. Kyrillidis, and V. Vassiliadis.
Improving Co-Cluster Quality with Application to Product Recommendations
ACM International Conference on Information and Knowledge Management, CIKM 2014
P. Bouboulis, S. Theodoridis, C. Mavroforakis, and L. Dalla.
Complex Support Vector Machines for Regression and Quaternary Classification
IEEE Transactions on Neural Networks and Learning Systems, TNNLS 2013
P. Bouboulis, S. Theodoridis, and C. Mavroforakis.
Complex Support Vector Regression
IEEE Workshop on Cognitive Information Processing, CIP 2012
N. Dimitropoulos, C. Mavroforakis, H. Georgiou, M. Mavroforakis, P. Bouboulis.
Original Open-Access Image Database of High-Quality Digital Mammographies for the Research Community (in Greek)
2nd Panhellenic Conference in Oncological Imaging (PCOI), 2011 [summary]
N. Dimitropoulos, H. Georgiou, M. Mavroforakis, P. Bouboulis, C. Mavroforakis.
MedSIP: Original Annotated Database of High-Quality Digital Mammographic Images Publicly Available to the Research Community (in Greek)
37th Annual Greek Medical Conference (EPIS), 2011
Setup and Administration of Lab's Computational Cluster
I am responsible for setting up and administrating a cluster of 14 machines that our lab acquired. The Cloudera distribution was installed on the machines and configured to be able to serve multiple users.
An extention to context tree weighting algorithms to problems involving categorical data and their relationship with Bayesian model selection (Athens University of Economics and Business, Greece)
During this project, I studied a family of algorithms known as "Context Tree Weighting" and I developed an efficient and fast implentation for it. After I completed my thesis, I worked on extending the theoretical framework to include categorical data and explored the application of this paradigm in Bayesian model selection. This work was funded by a "Basic Research (PEVE)" grant from AUEB, Greece.
Development of an annotated database of high-quality digital mammographic images
For this work, I closely collaborated with Dr. Nikolaos Dimitropoulos, MD, in order to gather high-quality (50μm, 12bit color, DICOM format) mammographic images from his "Senographe Essential" digital mammography system. The database was annotated using the BI-RADS mammography assesment protocol. This is the first free, online and fully annotated mammographic database at this quality level.
Using geometric Support Vector Machines to detect calcifications in heart CT scans (Computational Biomedicine Lab, UHouston)
In this project, I spent most of the time studying and implementing a new SVM algorithm, the geometric SVM. Following that, I employed this tool to detect and evaluate calcifications in the coronary arteries, using a large corpus of heart CT scans.